Night Sky Shooting / Astrophotography on X-T2 : ETTR? ISO 1600?

Sebastien Guyader wrote:

I've been playing around with Sequator, and I thought I'd share what seems to be the a workflow for the cleanest results.

1) Capture: set the camera to Raw, ISO to 800-3200, shutter speed 15-30sec (ISO and shutter speed will depend on the focal length and maximum aperture of your lens).

2) Raw processing: a minimal processing in a good raw converter is helpful. I use Rawtherapee, which allows me to apply white balance (important to me), vignetting and chromatic aberration and choosing the demosaicing algorithm, and to output linear 16-bit TIFF files (recommended input types for Sequator).

3) Aligning and stacking in Sequator:

• Composition: Align stars, Select best pixels (and if needed, Freeze ground)
• Reduce light pollution: Uneven
• Enhance star light: On
• Color space: Linear
• Output: uncompressed tiff

4) Postprocessing: I use Rawtherapee again for exposure compensation (to recover brightness) and final tuning using Lab adjustments (L* curve, chromaticity, fine tuning f color balance with a* or b* curve).

Below is a result I obtained form aligning and stacking 11 exposures (23mm @f/1.6, 15sec., ISO 2000) + 2 dark (noise) frames.

The result is much nicer to me than what I obtained using Siril for aligning and stacking (results I have posted some time ago here: https://www.dpreview.com/forums/post/61698294).

Thanks for sharing this & your previous pointers on RT. Looks like you nailed the focus - well done. Got questions & maybe a pointer or 2.

During RAW conversion for this in RT, did you apply any highlight reconstruction or compression or exposure compensation? I read the instructions for Sequator & disagree w/ their statement: "Any processing except vignetting correction should be avoided." Highlight recovery & several other adjustments are best done during RAW conversion.

Did you try a version of your stack w/ Reduce LP & Enhance Stars off? You have variable color (look at upper right vs lower left corner - lower left is almost devoid of color), perhaps too many blue stars & nearly uniform size for brighter stars. Hard to tell if these effects are from RT or Sequator.

Sequator Enhance stars may be brightening the brighter stars causing an increase in their size. The blue stars may be the result of your WB adjustments or uncorrected fringing & will limit how much color enhancement you can apply. The LP correction may be zapping color in parts of the image. There's more color in there.

Couple things you could try to bring back some of the sense of depth in the star field. Remember when you were looking at this area w/ your eyes, there was an obvious 3D sense to what you saw. That's been somewhat lost here. This is a big struggle for me too & many factors are involved.

Consider a couple things if you haven't done these. Look on YouTube for tutorials on how to reduce star size using PS. Try out applying RL Deconvolution sharpening to the final stack in RT.

How much have you cropped out of this image? The edges look much sharper than I'm getting from my XF23. Given the high astigmatism, its not easy to pick best focus for this one due to the astigmatism.

Last but not least, there are plenty of dark stellar dust trails in this region & they are almost completely lost in this image. They are really hard to tease out. Like nearly all Sony sensor cams, the Fuji Xs slightly clip blacks to reduce noise at high ISO. That's why shadow improvement as ISO is increased is so limited. Take a look at the shadow improvement curve for the Canon 6D.2.  The LP correction in Sequator may also clip blacks.

Several steps might help bring the dust trails out better. Bring up blacks a bit during RAW conversion so these aren't lost at the start. Instead of going for a completely black blank sky, aim for a luminosity of a few percent above black. Apply a contrast increase to shadows using your tool of choice to your stack. If you use curves to stretch the stack, make sure blacks are not being clipped further.

I may have questions for you about some detail about RT DCPs later in a PM.

Cheers,